PLANT DISEASE DIAGNOSIS - The Epidemiology of Plant Diseases - page 17

(Robinson, 1988) rather than fungal or bacterial plant pathogens as the banding
pattern becomes more complex as the genome becomes larger.
These techniques have been particularly popular for detecting unidentified
viruses or viroids in plants. This is important where plants are propagated vegeta-
tively or by seed (e.g. Robinson, 1988); unexpected pathogens cannot be detected by
specific antibodies as these must be raised in advance to familiar pathogens.
Methods based on nucleic acid differences are valuable as only these techniques can
be used to detect if any alien nucleic acid is present in plants that are being held in
quarantine. Discovering the presence of a foreign nucleic acid in addition to that of
the host plant can give warning of infection by any pathogen, albeit of some
unknown or symptomless disease. This diagnosis includes viroids that lack a protein
coat and so cannot be detected by antibodies. In this case, the identification of the
pathogen is of secondary importance since it is essential that the diagnostic test
should not be specific to any particular disease agent but be capable of detecting all
types of invading organisms. By constructing nucleic acid probes with various
breadths of specificity to pathogens, it is therefore possible to make probes that are
able to sort out the genetic relationships between them. Although a well chosen
collection of antibodies may also be capable of this, it is typically achieved only by
the chance selection of an appropriate antibody - hence success is generally more
elusive and laborious. For this reason, while it is often not feasible to select an anti-
body to detect a strain, a suitable individual nucleic acid probe is often capable of
detecting a range of strains (Robinson et al., 1987; Robinson and Legorburu, 1988).
Although methods based on differences in nucleic acid have so far been less widely
used in epidemiological studies than immunological techniques, they have proved
their value for determining the extent to which genets of Armillaria spp. have spread
and vegetatively mutated (Smith et al., 1992).
A multitude of methods to compare the genetic codes that were mostly pioneered
in other areas of biology and medicine have now been investigated for the diagnosis
of plant pathogens. Despite procedures as varied as melting nucleic acid to deter-
mine its mean content (Motta et al., 1986) and restriction site mapping (Anderson
et al., 1987), the most widely used techniques adopted involve the detection of
RFLP in the nucleic acid of pathogens by hybridization, versions of which have
become widely used to 'fingerprint' minor differences in genetic polymorphisms in
human populations.
Nucleic acid hybridization, or reassociation, is a process by which complemen-
tary single-stranded (ss) nucleic acid anneals to double-stranded (ds) nucleic acid.
This process occurs because of the hydrogen bonds between the two strands of the
DNA:DNA, DNA:RNA or RNA:RNA duplex and base stacking within the strands.
The subsequent avidity and the energy released during hybridization ensure the
stability of the hybrids (Tijssen, 1993). The homology of complementary base-pairs
ensures that pairing occurs with great specificity as adenine (A) pairs only with
thymine (T) - or uracil (U) in RNA - and guanine (G) with cytosine (C). For most
conventional hybridization tests, sensitivity of detection is comparable with ELISA
(Maule et al., 1983). A 17 base-pair (bp) oligonucleotide probe has been shown to
detect a single gene in a genome of 3 x 10 9 (Berent et al., 1985). Even gene
mutations of a single base-pair can be detected in pathogenic fungi (Martin et al.,